JULY 21, 2007 ORACLE ’S ACQUISITION OF PEOPLE SOFT : U.S. V. ORACLE R. PRESTON MCAFEE DAVID S. SIBLEY MICHAEL A. WILLIAMS Preston McAfee served as an economic expert on behalf of the U.S. Department of Justice, Antitrust Division in the matter of U.S. et al. v. Oracle . Michael Williams and the ERS Group were retained by the U.S. Department of Justice’s Antitrust Division to assist in the development of the economic analysis underlying McAfee’s testimony. During this period, David Sibley was Deputy Assistant Attorney General for Economics at the Antitrust Division. INTRODUCTION On June 6, 2003, the Oracle Corporation made an unsolicited cash tender offer for all the outstanding shares of PeopleSoft, Inc. Oracle and PeopleSoft are enterprise software companies that develop, manufacture, market, distribute, and service software products designed to assist businesses manage their operations. Together with SAP AG, they are the three largest companies in the industry. Oracle’s total revenues in fiscal year 2004 were $10.1 billion, while PeopleSoft and SAP AG had total revenues in 2003 of $2.3 billion and $8.0 billion, respectively. As discussed in detail below, all three firms produce enterprise resource planning (ERP) software that enables companies to operate their human resources, finances, supply chains, and customer relations. In February 2004, the U.S. Department of Justice (DOJ), together with the states of Connecticut, Hawaii, Maryland, Massachusetts, Michigan, Minnesota, New York, North Dakota, Ohio, and Texas (plaintiffs) filed suit to enjoin permanently Oracle’s acquisition of PeopleSoft. 1 The case raises several interesting antitrust issues. The DOJ’s case was founded on the theory that the merger would adversely affect ERP buyers because of unilateral competitive effects. 2 That is, the DOJ did not assert that the merger would lead to tacit or explicit collusion, but rather that the merger would cause the merging parties to cease competing, which would adversely affect customers for whom the two companies’ products were the first and second choices (holding constant rivals’ competitive strategies). In particular, the DOJ argued that buyers procured ERP 1 The February 26, 2004 complaint and other relevant legal documents and trial exhibits are available at http://www.usdoj.gov/atr/cases/oracle.htm. Since the DOJ was the lead party among the plaintiffs, we will frequently refer to the DOJ in representing the arguments of the plaintiffs. 2 U.S. Department of Justice & Federal Trade Commission, Horizontal Merger Guidelines (1992, rev. 1997), reprinted in 4 Trade Reg. Rep. (CCH) ¶ 13,104, at § 2.2 (hereinafter “Horizontal Merger Guidelines”). software in a manner appropriately modeled by auction theory, and the proposed merger would eliminate PeopleSoft as a bidder. As part of its analysis, one of the DOJ’s experts presented a merger simulation model based on auction theory. United States v. Oracle represents the first case in which a merger simulation was used in court. What should be the role of merger simulations in analyzing the competitive effects of horizontal mergers? In particular, what should be the role of market definition in merger simulations of unilateral effects cases? The DOJ and Oracle also relied extensively on customer testimony in providing evidence pertaining to market definition and the likely competitive effects of the proposed merger. What should be the role of customer testimony in analyzing these two issues? After a trial in U.S. District Court for the Northern District of California, Judge Vaughn Walker ruled on September 9, 2004 that the DOJ had not proven its case that the proposed merger would violate U.S. antitrust law. The DOJ announced on October 1, 2004, that it would not appeal Judge Walker’s decision. In December 2004, PeopleSoft’s Board of Directors accepted Oracle’s $10.3 billion offer. MARKET BACKGROUND : ENTERPRISE APPLICATION SOFTWARE The proposed merger of Oracle and PeopleSoft raised concerns about software products belonging to the broad category commonly called “enterprise application software” (EAS), of which ERP is one type. These products are used to automate the performance of necessary business functions. Important segments (or “pillars”) include the following: (1) “Human resources management” (HRM) software, which automates payroll services, recruiting, training, and benefits administration; (2) “Financial management systems” 2 (FMS) software, which automates the general ledger, accounts payable and receivable, and asset management; (3) “Supply chain management” (SCM) software, which assists in the control of inventory, manufacturing, and distribution; and (4) “Customer relations management” (CRM) software, which manages the entire life cycle of a sale, from the development of customer prospects to customer support and service. Each pillar may contain from 30 to 70 modules. Some companies sell suites containing modules from more than one pillar. Such combinations of pillars are referred to as ERP suites. Typically, an ERP suite is a collection of packaged software that integrates most of a firm’s data across most of its activities. When an individual pillar is sold on a stand- alone basis, it is known as a “point solution” or “best-of-breed solution.” Businesses vary greatly in size, complexity, and the efficiency of the embedded or legacy information technology infrastructure used to support their operations. As a result, businesses vary greatly with respect to the features that they value in ERP and their willingness to pay for these features. A product that meets the requirements of one large, multinational corporation with global operations may not meet the requirements of another large, multinational corporation or the requirements of a small, single- establishment business. For these reasons, ERP software exhibits considerable product differentiation, with vendors often focusing on the requirements of specific industries (or “verticals”) such as banking, healthcare, and government. Vendors also differ with respect to the types of software that they develop, with some firms focusing on off-the- shelf products that serve the requirements of smaller firms with relatively simple business operations, and with others focusing on complex, customizable software and ancillary services designed to meet the specific requirements of large, complex enterprises 3 (LCEs). 3 For LCEs, the fees to license and maintain ERP software are often only 10% to 15% of the total cost of ownership, which also includes the costs of personnel training, consulting, and integrating the new program with the customer’s legacy software and databases. Large firms that purchase complex, customizable software typically rely on competitive bids to procure solutions to their business requirements. Such firms identify a relatively small number of vendors that are capable of meeting their requirements, send them requests for proposals (RFPs), and engage in protracted negotiations with the vendor(s) that submit the most attractive initial responses. THE DOJ’S CASE Competitive Effects Theory The DOJ began by distinguishing three categories of EAS: (1) “off-the-shelf” PC-based products suitable for many small businesses; (2) relatively inexpensive software with limited capability that must be professionally installed and maintained and is suitable for “mid-market” firms; and (3) “high-function enterprise software” required by “enterprise customers” or LCEs. High-function products support thousands of simultaneous users and tens of thousands of simultaneous transactions, integrate seamlessly across pillars (specifically, HRM and FMS modules), and are sufficiently flexible to support the unique business processes of each LCE. 3 “Customizable software” means that the software can be configured to fit a customer’s requirements. This does not mean that the fundamental software code is modified for a given customer. The software that Oracle, for example, sells to a large, multinational company is identical to the software it sells to a small company. “Customization” is achieved through the use of software settings that can be configured to match a customer’s requirements. 4 Critical features of high-function enterprise software include the ability to support business operations that span (1) multiple jurisdictions with multiple currencies and languages; (2) multiple legal entities or divisions within the business; and (3) multiple lines of business. High-function enterprise software is also distinguished by (1) its total cost of ownership, which amounts to millions of dollars; (2) the relatively long period (e.g., several months) taken by customers to make a purchase decision; and (3) the difficulty in implementing the software. Enterprise customers will not consider vendors that cannot provide continuous technical support and continuous enhancements to the product’s capabilities over its long life. LCEs are also unwilling to consider vendors that lack a track record of successful implementation of high-function enterprise software suites. The typical procurement cycle of large, complex enterprises includes the following steps: the enterprise analyzes and identifies its requirements; determines the expected return on investment and prepares a budget; forms a selection committee; establishes detailed functional requirements; issues a Request for Information (RFI) to pre-screen possible vendors; issues RFPs to qualified vendors; arranges for demonstrations from three to five qualified vendors; receives bids;